Use Case

We currently only need a (big) subset from one of our edge devices, but this results in Telegraf sending about 200 read requests which takes about 4 seconds. The subsets are not completely contiguous in the address space, but the gaps in between the blocks are only about 1-4 registers. If we manually fill these gaps with "Spare" fields the performance increase significantly to 150ms.

We then manually drop the Spare* fields.

Expected behavior

Telegraf sends one big read request for the entire configured register span and drops the unused fields.

Actual behavior

Separate read requests are sent for every contiguous register block, which slows down performance a lot.

Additional info

No response